Methods and systems for transmitting media streams over variable bandwidth channels, are known in the art.
U.S. Pat. No. 5,847,760 issued to Elmaliach et al. and entitled “Method for Managing Video Broadcast”, is directed to a method for broadcasting compressed digital video signals. An encoder encodes each video frame according to the maximal bit rate, the half-maximal bit rate, and the quarter of the maximal bit rate. Each frame contains a plurality of a group of pictures (GOP). Each frame contains the same number of GOP's, and the GOP's are placed in the same order in all three types of frames. A transmitter detects the available bandwidth in the communication link, and provides this information to a controller. The controller selects one of the three types of encoded frames. If the available bandwidth is maximal, the controller selects the maximal bit rate frame type. If the available bandwidth is less than the maximal but more than half, the controller selects the half-maximal frame type, and if the available bandwidth is less than half the maximal, then the controller selects the quarter bit rate frame. A receiver receives the compressed frame from the transmitter and forwards it to a decoder, which decodes the frame and displays it on a display.
Methods for multiplexing a plurality of media streams into a single channel, where each media stream is assigned a different bandwidth, are also known in the art. One such method analyzes the activity in the streams and provides feedback to the encoders to encode according to a selected bit rate.
The Rate-Mux 300, Cisco V-bits Corp. implements another method in which the unit's programmable statistical re-multiplexing engine performs real-time CBR (Constant Bit Rate) and VBR (Variable Bit Rate) rate conversion and bit-stream manipulation processing to the input streams. Using a statistical model, the processed streams are then recombined into a highly efficient, multiplexed program transport stream for retransmission.
U.S. Pat. No. 5,966,120 to Arazi et al, is directed to “Method and apparatus for combining and distributing data with pre-formatted real-time video”. This reference describes a system, which transmits a plurality of streams via a constant bit rate channel. Each of the streams can be of a variable bit rate channel but the system assures that the total sum of all if the streams at any given point in time does not exceed the channel bandwidth, even if from time to time, not all of it is fully used. The system described in this reference attempts to make use of excess bandwidth, which is variably available and insert other types of data therein, such as commercials. This system requires a client unit at the receiving end to extract the transmitted streams and the embedded additional data.
U.S. Pat. No. 5,629,736 issued to Haskell et al. and entitled “Coded Domain Picture Composition for Multimedia Communications Systems”, is directed to a method for composition of video signals. Up to four video signals from four sources are fed to the system. The bit rates of the four signals may be different. Each video signal is decoded and processed by a discrete cosine transformation processor. The four discrete cosine transformation processors are controlled by a controller, whereby the bit rates of all four signals are converted to a single bit rate. The four video signals of equal bit rates are then synchronized and multiplexed, and fed to a display unit as one video signal, to display the four video images simultaneously.
U.S. Pat. No. 5,216,503 issued to Paik, et al. and entitled “Statistical Multiplexer for a Multichannel Image Compression System”, is directed to a system for compression of a plurality of video signals and decompression thereof. Each of a plurality of video signals is encoded by an encoder. The encoded video signals are multiplexed and transmitted as a single video signal to a receiver. The video signal is then demultiplexed and decoded by a number of decoders, each corresponding to an encoder at the transmitting end. Each decoder then outputs the original video signal encoded at the transmission end.
MPEG standard provides a dual layer method of encoding, which is called SNR scalability. According to this method, a stream is encoded into two encoded streams, a basic level stream and a supplemental enhanced level stream. The layering is performed once for the entire movie and not on a block level. Either the base layer or both layers can be used to produce the original stream. The layer selection decision has to be performed prior to transmission.